15 results about "Classification types" patented technology

The invention relates to the technical field of intelligent recommendation, and discloses a hybrid retrieval RAG-based resource recommendation method and system, and the method comprises the steps: collecting resource text data, and constructing a vector library and a tag library; expanding the user question based on the language model to obtain a plurality of semantic extension questions; performing intention recognition, judging whether the user question is a resource recommendation question, and if yes, determining a target classification type; screening the data according to the field definition in the tag library to obtain a candidate knowledge fragment set; obtaining candidate vectors, mapping the user question and the semantic extension question into query vectors, calculating the similarity between the query vectors and each candidate vector, and selecting knowledge supplement content; and performing resource splicing on all the knowledge supplement contents to generate resource recommendation answers. According to the method, a structured label screening mechanism and a semantic vector fine arrangement mechanism are fused, and the problems of recall redundancy, matching deviation and the like caused by the fact that an existing RAG system only depends on semantic similarity retrieval are solved.

This application relates to a data transmission method, apparatus, computer equipment, computer-readable storage medium, and computer program product, belonging to the field of data transmission technology, and can be used in the financial technology field or other related fields. The method includes: classifying the business data to be transmitted to obtain the corresponding classification type; when the classification type is a newly added type, determining the target compression strategy corresponding to the classification type from a plurality of preset compression strategies, and transmitting the business data under the classification type based on the target compression strategy; when the classification type is an existing type, transmitting the business data under the classification type based on the target compression strategy. This method can improve the utilization rate of bandwidth resources.

The utility model discloses a classification type living body specimen storage device, which relates to the technical field of living body specimen storage and comprises a culture box body, a box door is mounted on the front side of the culture box body, side plates are symmetrically connected to the inner wall of the culture box body, and classification plates are arranged on the sides, close to each other, of the side plates. By arranging the limiting grid plate and the vertical rods, in the using process, the limiting grid plate and the vertical rods are matched with each other, so that a worker can conveniently limit and fix a container placed above the classification plate, and the situation that the container falls off in the drawing process of the classification plate is effectively prevented; the safety of the device in the using process can be effectively guaranteed, a connecting strip and a sliding block are arranged, when a container borne above a classification plate is taken, a worker can transversely push a limiting grid plate, so that the limiting grid plate and the classification plate are staggered, the space for the worker to take the container can be increased, and the working efficiency is improved. Therefore, the convenience of the device in the using process can be effectively improved.

The application provides a commodity comment sentiment tendency analysis method, device, equipment and readable storage medium, wherein a keyword contained in a to-be-processed comment is acquired, the keyword is one or more of preset target feature words, and a classification type word best reflecting a to-be-processed comment is obtained; a sentiment feature vector corresponding to the to-be-processed comment is acquired according to the keyword; a sentiment feature vector of the to-be-processed comment is processed by using a preset sentiment tendency analysis model, and a sentiment type of the to-be-processed comment is acquired, so that the accuracy of sentiment tendency analysis is improved.

The application relates to the technical field of intelligent recommendation, and discloses a resource recommendation method and system based on a hybrid retrieval RAG, which comprises the following steps: collecting resource text data, constructing a vector library and a label library; expanding a user question based on a language model to obtain a plurality of semantic expansion questions; performing intention recognition to determine whether the user question is a resource recommendation question, and determining a target classification type if the user question is a resource recommendation question; filtering data according to field definitions in the label library to obtain a candidate knowledge segment set; obtaining candidate vectors, mapping the user question and the semantic expansion questions into query vectors, calculating the similarity between the query vectors and each candidate vector, and selecting knowledge supplement content; and performing resource splicing on all the knowledge supplement content to generate a resource recommendation answer. The application combines a structured label filtering mechanism and a semantic vector precision sorting mechanism, and overcomes the problems of recall redundancy and matching deviation caused by the fact that the existing RAG system only depends on semantic similarity retrieval.

The utility model relates to the technical field of electromechanical machinery, in particular to a classification type double-disc take-up machine take-up mechanism which comprises a base, an adjusting assembly, a driving assembly, a take-up assembly and a flattening assembly. Mounting grooves are formed in the two sides of the base, a moving groove is formed in the upper portion of the base, the mounting grooves are connected with mounting rods, and mounting holes a are formed in the mounting rods; the adjusting assembly is connected to the base; the adjusting assembly comprises an adjusting plate which is connected with a bottom plate, and the bottom plate is connected with moving wheels; the two sides of the bottom plate are connected with vertical plates, and the two sides of the vertical plates are connected with electric telescopic rods; the adjusting plate is connected with a reinforcing piece; the driving assembly is mounted on the adjusting assembly; the take-up assembly is mounted on the driving assembly; the take-up assembly comprises a take-up roller, the peripheries of the two ends of the take-up roller are connected with fixing plates, the fixing plates are connected with extension plates, and the extension plates are connected with connecting plates. The distance of the classification type double-disc take-up mechanism can be conveniently adjusted, the adjusting assembly is operated, the take-up range is conveniently widened, and the working efficiency is improved.

The present invention accurately classifies a sentence if sentences are classified by using a generative AI model.　A sentence classification device 10 comprises: an acquisition unit 11 that acquires information indicating a sentence to be classified, a plurality of sentence examples divided into a plurality of groups, and an example of classification of each sentence example into any of a plurality of classification types; a model usage unit 12 that acquires, for each group and from a generative AI model 21, a classification result for the sentence to be classified, by inputting, into the generative AI model 21, the sentence to be classified, a sentence example included in the group, and the classification example for each sentence example, which were acquired by the acquisition unit 11; and a determination unit 13 that determines a final classification of the sentence to be classified from the classification result for each group acquired by the model usage unit 12.

The present disclosure provides a question analysis method and device, electronic equipment and storage medium, relates to the technical field of Internet, and particularly relates to the field of knowledge graph. The specific implementation scheme is: obtaining a question input by a user; performing knowledge classification on the question according to a pre-constructed classification type to obtain classification information of the question; performing knowledge understanding on the question to obtain subject segment understanding information and predicate segment understanding information of the question; and generating final analysis information of the question based on the subject segment understanding information and the predicate segment understanding information and the classification information. The embodiments of the present application can provide a general analysis scheme for knowledge class Query, can be applied in different scenes with knowledge class demand, and have stronger explainability and higher accuracy.

The invention discloses a semantic retrieval method and device. The method comprises the steps that retrieval content is input into an atomic condition recognition model to generate an atomic condition information table; inputting the retrieval content semantic information into an external general large model to obtain an expression pseudo code; inputting the atomic conditions, the retrieval content and the expression pseudo code into an external general large model to obtain atomic conditions with classification types, and extracting and assembling according to semantic information to obtain a relationship among the atomic conditions; extracting retrieval elements from the semantic information of the retrieval content; classifying the expanded entity elements to obtain classification limiting conditions, and fusing the classification limiting conditions into an atomic condition information table to obtain an updated atomic condition relationship again; according to different application requirements and the expanded retrieval entity elements, a triple model, a quadruple model or a multi-tuple model is established to form a vector model; and according to a relationship between the vector model and the updated atomic condition, performing fusion retrieval to obtain a retrieval result. According to the method, semantic retrieval can be efficiently and accurately carried out.

The invention relates to the technical field of video intelligent classification, and discloses a video object AI classification processing system fused with MPEG-4 coding. The system comprises an analysis module, a feature extraction module, a classification model module, a fusion module and an updating module. The analysis module analyzes the MPEG-4 coded data, extracts video object information containing multiple features, and generates dynamic priority parameters; the feature extraction module extracts multi-modal features and generates a feature vector set; the classification model module calls the pre-training network to process the feature vector set and outputs a classification result; the fusion module fuses the multi-source results to generate a target classification type based on the dynamic priority parameters and the historical classification record reference mode; and the updating module optimizes the classification path according to the classification conflict index and corrects network parameters. The system can make full use of multi-feature information, improves the classification accuracy, flexibility and reliability, and can effectively meet the classification requirements of complex video scenes.

A method of improving performance of classification on the basis of mixed sampling is applied. The present invention is directed to providing a method and apparatus for improving the performance of classification on the basis of mixed sampling that are capable of, when learning a model that classifies types using deep learning by dividing the entire data set and using the divided data set for training, validation, and testing, applying different sampling techniques by types of data according to the characteristics of the training data in order to improve the classification performance.

The invention relates to the technical field of chip verification, in particular to a verification platform configuration method, which comprises the following steps: acquiring N hardware input ports related to a current verification target in a tested design; obtaining a preset classification strategy of the verification platform and M interface sets configured according to the classification strategy; dividing the N hardware input ports into M port classification sets according to a classification strategy, and binding and connecting the hardware input ports in the M port classification sets and the interfaces in the M interface sets of the verification platform; binding an independent special input agent for each interface set in the verification platform; when the verification sequence is started each time, the transaction generated by the verification sequence is obtained, according to the classification type of data in the transaction, the verification sequence is connected to the sequencer of the special agent matched with the classification type, then excitation is applied to the hardware input port through the driver so as to drive the tested design, and universality of the verification platform is achieved.

This specification relates to the field of computer technology and provides a method and apparatus for merging and processing unstructured data. The method includes: acquiring multiple unstructured data sets to be processed; flattening the multiple unstructured data sets using preset data identifiers to obtain multiple flattening results; comparing the multiple flattening results to obtain difference information between them; classifying the difference information to obtain classification types; and merging the multiple flattening results using a preset processing strategy corresponding to the classification types, with a preset flattening result among the multiple flattening results as the standard. Through the embodiments of this specification, refined merging of unstructured data can be achieved.

This application discloses a training method, intrusion detection method, and apparatus for an intrusion detection model. It involves collecting production process data from an industrial control system, using this data as training data. Feature extraction is performed on the production process data to obtain production process feature data. This feature data, along with random noise, is input into a generative adversarial network (GAN) to obtain adversarial example data. To further improve the quality of the adversarial example data, it undergoes weighting and random projection processing to obtain target adversarial example data. This target adversarial example data is then input into the intrusion detection model to obtain its predicted classification type. Subsequently, the intrusion detection model can be trained based on the predicted classification type and type label of the target adversarial example data. Thus, training the intrusion detection model with high-quality adversarial example data improves its robustness and accuracy.

The invention provides a policy intelligent keyword generation method which comprises the following steps: S1, monitoring a policy release source target website in real time through technologies such as crawlers and the like, and timely collecting the latest webpage content to the local to finish local storage of the policy content; s2, performing large classification on the whole policy, then performing fine classification on the text content of the policy, and extracting policy declaration requirements; s3, carrying out keyword extraction on the main body content of the policy through TEXT-RANK and TF-IDF algorithms on the main body content of the policy subjected to type segmentation; and S4, the first-level classification of the policy and the keywords of the policy content are displayed, and the enterprise can quickly match with the policy based on the types and the keywords. By classifying the policies and generating the policy content keywords, policy propaganda and better policy interpretation of an enterprise terminal are facilitated, accurate policy matching is realized, and policy landing efficiency and convenience for enterprises to obtain policy support are improved.